Simultaneous truth and performance level estimation through fusion of probabilistic segmentations

Automatic tissue segmentation of head and neck MR images for hyperthermia treatment planning

A hyperthermia treatment requires accurate, patient-specific treatment planning. This planning is based on 3D anatomical models which are generally derived from computed tomography. Because of its superior soft tissue contrast, magnetic resonance imaging (MRI) information can be introduced to improve the quality of these 3D patient models and therefore the treatment planning itself. Thus, we present here an automatic atlas-based segmentation algorithm for MR images of the head and neck. Our method combines multiatlas local weighting fusion with intensity modelling. The accuracy of the method was evaluated using a leave-one-out cross validation experiment over a set of 11 patients for which manual delineation were available. The accuracy of the proposed method was high both in terms of the Dice similarity coefficient (DSC) and the 95th percentile Hausdorff surface distance (HSD) with median DSC higher than 0.8 for all tissues except sclera. For all tissues, except the spine tissues, the accuracy was approaching the interobserver agreement/variability both in terms of DSC and HSD. The positive effect of adding the intensity modelling to the multiatlas fusion decreased when a more accurate atlas fusion method was used.Using the proposed approach we improved the performance of the approach previously presented for H&N hyperthermia treatment planning, making the method suitable for clinical application.

Towards measuring neuroimage misalignment

To enhance neuro-navigation, high quality pre-operative images must be registered onto intra-operative configuration of the brain. Therefore evaluation of the degree to which structures may remain misaligned after registration is critically important. We consider two Hausdorff Distance (HD)-based evaluation approaches: the edge-based HD (EBHD) metric and the Robust HD (RHD) metric as well as various commonly used intensity-based similarity metrics such as Mutual Information (MI), Normalised Mutual Information (NMI), Entropy Correlation Coefficient (ECC), Kullback-Leibler Distance (KLD) and Correlation Ratio (CR). We conducted the evaluation by applying known deformations to simple sample images and real cases of brain shift. We conclude that the intensity-based similarity metrics such as MI, NMI, ECC, KLD and CR do not correlate well with actual alignment errors, and hence are not useful for assessing misalignment. On the contrary, the EBHD and the RHD metrics correlated well with actual alignment errors; however, they have been found to underestimate the actual misalignment. We also note that it is beneficial to present HD results as a percentile-HD curve rather than a single number such as the 95-percentile HD. Percentile-HD curves present the full range of alignment errors and also facilitate the comparison of results obtained using different approaches. Furthermore, the qualities that should be possessed by an ideal evaluation metric were highlighted. Future studies could focus on developing such an evaluation metric.

A Model of Population and Subject (MOPS) Intensities With Application to Multiple Sclerosis Lesion Segmentation

White matter (WM) lesions are thought to play an important role in multiple sclerosis (MS) disease burden. Recent work in the automated segmentation of white matter lesions from magnetic resonance imaging has utilized a model in which lesions are outliers in the distribution of tissue signal intensities across the entire brain of each patient. However, the sensitivity and specificity of lesion detection and segmentation with these approaches have been inadequate. In our analysis, we determined this is due to the substantial overlap between the whole brain signal intensity distribution of lesions and normal tissue. Inspired by the ability of experts to detect lesions based on their local signal intensity characteristics, we propose a new algorithm that achieves lesion and brain tissue segmentation through simultaneous estimation of a spatially global within-the-subject intensity distribution and a spatially local intensity distribution derived from a healthy reference population. We demonstrate that MS lesions can be segmented as outliers from this intensity model of population and subject. We carried out extensive experiments with both synthetic and clinical data, and compared the performance of our new algorithm to those of state-of-the art techniques. We found this new approach leads to a substantial improvement in the sensitivity and specificity of lesion detection and segmentation.

Optimal MAP Parameters Estimation in STAPLE Using Local Intensity Similarity Information

In recent years, fusing segmentation results obtained based on multiple template images has become a standard practice in many medical imaging applications. Such multiple-templates-based methods are found to provide more reliable and accurate segmentations than the single-template-based methods. In this paper, we present a new approach for learning prior knowledge about the performance parameters of template images using the local intensity similarity information; we also propose a methodology to incorporate that prior knowledge through the estimation of the optimal MAP parameters. The proposed method is evaluated in the context of segmentation of structures in the brain magnetic resonance images by comparing our results with some of the state-of-the-art segmentation methods. These experiments have clearly demonstrated the advantages of learning and incorporating prior knowledge about the performance parameters using the proposed method.

Voxel-based dipole orientation constraints for distributed current estimation

Distributed electroencephalography source localization is a highly ill-posed problem. With measurements on the order of 10(2), and unknowns in the range of 10(4)-10(5), the range of feasible solutions is quite large. One approach to reducing ill-posedness is to intelligently reduce the number of unknowns. Restricting solutions to gray matter is one approach. A further step is to use the anatomy of each patient to identify and constrain the orientation of the dipole within each voxel. While dipole orientation constraints for cortical patch-based approaches have been proposed, to our knowledge, no solutions for full volumetric localizations have been presented. Patch techniques account for patch surface area, but place dipoles only on the surface, rather than throughout the cortex. Variability in human cortical thickness means that thicker regions of cortex will potentially contribute more to the EEG signal, and should be accounted for in modeling. Additionally, patch models require cortical surface identification techniques, which can separate them from the extensive literature on voxel-based MR image processing, and require additional adaptation to incorporate more complex information. We present a volumetric approach for computing voxel-based distributed estimates of cortical activity with constrained dipole orientations. Using a tissue thickness estimation approach, we obtain estimates of the cortical surface normal at each voxel. These let us constrain the inverse problem, and yield localizations with reduced spatial blurring and better identification of signal magnitude within the cortex. This is demonstrated for a series of simulated and experimental data using patient-specific bioelectric models.

An algorithm for optimal fusion of atlases with different labeling protocols

In this paper we present a novel label fusion algorithm suited for scenarios in which different manual delineation protocols with potentially disparate structures have been used to annotate the training scans (hereafter referred to as "atlases"). Such scenarios arise when atlases have missing structures, when they have been labeled with different levels of detail, or when they have been taken from different heterogeneous databases. The proposed algorithm can be used to automatically label a novel scan with any of the protocols from the training data. Further, it enables us to generate new labels that are not present in any delineation protocol by defining intersections on the underling labels. We first use probabilistic models of label fusion to generalize three popular label fusion techniques to the multi-protocol setting: majority voting, semi-locally weighted voting and STAPLE. Then, we identify some shortcomings of the generalized methods, namely the inability to produce meaningful posterior probabilities for the different labels (majority voting, semi-locally weighted voting) and to exploit the similarities between the atlases (all three methods). Finally, we propose a novel generative label fusion model that can overcome these drawbacks. We use the proposed method to combine four brain MRI datasets labeled with different protocols (with a total of 102 unique labeled structures) to produce segmentations of 148 brain regions. Using cross-validation, we show that the proposed algorithm outperforms the generalizations of majority voting, semi-locally weighted voting and STAPLE (mean Dice score 83%, vs. 77%, 80% and 79%, respectively). We also evaluated the proposed algorithm in an aging study, successfully reproducing some well-known results in cortical and subcortical structures.

A logarithmic opinion pool based STAPLE algorithm for the fusion of segmentations with associated reliability weights

Pelvic floor dysfunction is common in women after childbirth and precise segmentation of magnetic resonance images (MRI) of the pelvic floor may facilitate diagnosis and treatment of patients. However, because of the complexity of its structures, manual segmentation of the pelvic floor is challenging and suffers from high inter and intra-rater variability of expert raters. Multiple template fusion algorithms are promising segmentation techniques for these types of applications, but they have been limited by imperfections in the alignment of templates to the target, and by template segmentation errors. A number of algorithms sought to improve segmentation performance by combining image intensities and template labels as two independent sources of information, carrying out fusion through local intensity weighted voting schemes. This class of approach is a form of linear opinion pooling, and achieves unsatisfactory performance for this application. We hypothesized that better decision fusion could be achieved by assessing the contribution of each template in comparison to a reference standard segmentation of the target image and developed a novel segmentation algorithm to enable automatic segmentation of MRI of the female pelvic floor. The algorithm achieves high performance by estimating and compensating for both imperfect registration of the templates to the target image and template segmentation inaccuracies. A local image similarity measure is used to infer a local reliability weight, which contributes to the fusion through a novel logarithmic opinion pooling. We evaluated our new algorithm in comparison to nine state-of-the-art segmentation methods and demonstrated our algorithm achieves the highest performance.

Hierarchical performance estimation in the statistical label fusion framework

Label fusion is a critical step in many image segmentation frameworks (e.g., multi-atlas segmentation) as it provides a mechanism for generalizing a collection of labeled examples into a single estimate of the underlying segmentation. In the multi-label case, typical label fusion algorithms treat all labels equally - fully neglecting the known, yet complex, anatomical relationships exhibited in the data. To address this problem, we propose a generalized statistical fusion framework using hierarchical models of rater performance. Building on the seminal work in statistical fusion, we reformulate the traditional rater performance model from a multi-tiered hierarchical perspective. The proposed approach provides a natural framework for leveraging known anatomical relationships and accurately modeling the types of errors that raters (or atlases) make within a hierarchically consistent formulation. Herein, the primary contributions of this manuscript are: (1) we provide a theoretical advancement to the statistical fusion framework that enables the simultaneous estimation of multiple (hierarchical) confusion matrices for each rater, (2) we highlight the amenability of the proposed hierarchical formulation to many of the state-of-the-art advancements to the statistical fusion framework, and (3) we demonstrate statistically significant improvement on both simulated and empirical data. Specifically, both theoretically and empirically, we show that the proposed hierarchical performance model provides substantial and significant accuracy benefits when applied to two disparate multi-atlas segmentation tasks: (1) 133 label whole-brain anatomy on structural MR, and (2) orbital anatomy on CT.

Electrode localization for planning surgical resection of the epileptogenic zone in pediatric epilepsy

PURPOSE

In planning for a potentially curative resection of the epileptogenic zone in patients with pediatric epilepsy, invasive monitoring with intracranial EEG is often used to localize the seizure onset zone and eloquent cortex. A precise understanding of the location of subdural strip and grid electrodes on the brain surface, and of depth electrodes in the brain in relationship to eloquent areas is expected to facilitate pre-surgical planning.

METHODS

We developed a novel algorithm for the alignment of intracranial electrodes, extracted from post-operative CT, with pre-operative MRI. Our goal was to develop a method of achieving highly accurate localization of subdural and depth electrodes, in order to facilitate surgical planning. Specifically, we created a patient-specific 3D geometric model of the cortical surface from automatic segmentation of a pre-operative MRI, automatically segmented electrodes from post-operative CT, and projected each set of electrodes onto the brain surface after alignment of the CT to the MRI. Also, we produced critical visualization of anatomical landmarks, e.g., vasculature, gyri, sulci, lesions, or eloquent cortical areas, which enables the epilepsy surgery team to accurately estimate the distance between the electrodes and the anatomical landmarks, which might help for better assessment of risks and benefits of surgical resection.

RESULTS

Electrode localization accuracy was measured using knowledge of the position of placement from 2D intra-operative photographs in ten consecutive subjects who underwent intracranial EEG for pediatric epilepsy. Average spatial accuracy of localization was 1.31 ± 0.69 mm for all 385 visible electrodes in the photos.

CONCLUSIONS

In comparison with previously reported approaches, our algorithm is able to achieve more accurate alignment of strip and grid electrodes with minimal user input. Unlike manual alignment procedures, our algorithm achieves excellent alignment without time-consuming and difficult judgements from an operator.